The Invoice After the Invoice: Build vs Buy and the UK Integration Tax in 2026

Every build vs buy debate starts with a tidy spreadsheet. On one side: licence fees, onboarding and a go-live date. On the other: sprint estimates, hiring and a roadmap. Then the real invoice arrives — not for the software, but for making it behave like a UK business app.

That gap is the integration tax. It is the cost of connectors, compliance evidence, data mapping, testing, support and, eventually, leaving. For UK mid-market teams in 2026, it is often the difference between a sensible buy and an expensive trap.

The demo never shows the integration queue

SaaS demos are polished because they show the happy path. Your finance team sees a clean dashboard. Your operations lead sees a workflow builder. Nobody sees the queue of work needed to connect that tool to HMRC, Sage, Xero, your identity provider, your data warehouse and whatever legacy system still runs the warehouse rota.

Buying does not remove that work. It moves it. Instead of building features, you build adapters, reconciliation jobs and evidence packs. The question is whether those costs are visible before signature.

HMRC Making Tax Digital is not a checkbox

Making Tax Digital for VAT is established, and MTD for Income Tax continues to reshape how sole traders and landlords keep records. If your product touches tax data, 'MTD-ready' on a vendor slide is only the start.

You still need to know how authentication works, how submissions are retried, how errors are surfaced to users and who owns the audit trail when a filing fails at 11pm. A bought tool may have HMRC recognition, but your workflows, permissions and support model still need to line up.

Common integration tax here: testing sandbox versus production behaviour, handling agent authorisations, and training support staff to diagnose rejections without blaming HMRC.

Sage and Xero connections are deeper than a logo

Many UK mid-market firms run finance on Sage or Xero. A marketplace listing that says 'integrates with Xero' can mean anything from a nightly CSV export to a bi-directional sync with granular permissions.

Ask the unglamorous questions. Which entities sync? What happens to credit notes? How are VAT codes mapped? What are the API rate limits? Who gets alerted when a sync fails halfway through a payroll run?

If finance data is central to your product, a shallow connector becomes a permanent support burden. Building your own connector is not free either — APIs change, scopes tighten and reconciliation logic grows. The decision hinges on how differentiated your finance workflow really is.

NHS login and health-adjacent products

If your app touches health data or claims to support NHS-facing journeys, generic customer identity tools may not be enough. NHS login has its own identity proofing, scopes and assurance expectations. Clinical safety, data security and DSPT evidence can enter the conversation earlier than product teams expect.

Buying a CIAM platform can help with authentication mechanics, but it does not automatically solve NHS-specific onboarding, consent or audit requirements. For a mid-market team, the integration tax may include a specialist partner, extra testing and a longer security review than the original business case assumed.

Data residency and UK GDPR: the cost of proving where things live

UK GDPR obligations do not disappear because a vendor has a UK region. You still need a data processing agreement, a record of processing, retention rules, breach processes and a clear view of sub-processors. International transfers need the right safeguards, and support access from outside the UK can complicate the story.

Off-the-shelf tools often store data in the UK but route support, analytics or backups elsewhere. Building gives you more control, but you inherit the duty to implement it properly. Either way, 'we take security seriously' is not evidence.

The hidden cost is usually documentation and review time: security questionnaires, ICO-facing paperwork, customer due diligence and the legal review of exit and deletion terms.

Vendor lock-in has a leaving tax

Lock-in is not just about pricing. It is about custom fields that do not export cleanly, workflows that only exist inside the vendor, marketplace add-ons that vanish, and API limits that make migration expensive. The leaving tax is real, and it should sit in the same spreadsheet as the licence fee.

For UK mid-market teams, the risk is sharper because headcount is lean. A two-person product team cannot absorb a six-month re-platforming project without pausing the roadmap.

A practical decision framework

Score each option from 1 to 5 on these weighted criteria, then discuss the outliers rather than trusting the total.

Criterion	What to ask	Weight
Regulatory criticality	Does this touch HMRC, health data, payments or regulated records?	High
Integration depth	Is the connector read-only, bi-directional or workflow-critical?	High
Data sensitivity	Would a residency or sub-processor surprise hurt customer trust?	High
Process differentiation	Is this workflow a reason customers choose you?	Medium
Volume and scale	Will API limits or seat pricing break at 3x growth?	Medium
Time to market	Can you afford a build cycle before revenue?	Medium
Exit cost	What does migration look like in 24 months?	High

A buy signal is strong when the workflow is commodity, the vendor has proven UK integrations and the exit path is clean. A build signal is stronger when the integration is the product, the data model is your moat or regulatory evidence must be owned in-house.

Five questions to ask before signing

1. Which UK-specific integrations are included, and which are professional services?
2. Can we see a real data mapping for Sage, Xero or HMRC — not a sales diagram?
3. Where is data stored, processed and supported, and which sub-processors are involved?
4. What does export look like, including custom fields and audit history?
5. Who is on the hook when a sync fails during month-end?

The bottom line

Build vs buy is not a binary. It is a bet on where you want to carry complexity. Buying moves complexity into contracts, connectors and compliance reviews. Building moves it into engineering, maintenance and on-call. For UK mid-market apps, the winning move in 2026 is to price both honestly — including the invoice after the invoice.